Sigmabrogz/agent-devtools

AI Summary: This issue requests the addition of cost tracking for new AI models, specifically Claude 3.5 Haiku and GPT-4.5, to the `MODEL_COSTS` dictionary in `config.py`. The implementation involves a straightforward update to this configuration file, making it an ideal task for new contributors.

AI Summary: This issue proposes enhancing error messages within the codebase to be more informative and actionable for common problems like WebSocket connection failures, lack of trace capture, and instrumentor wrapping issues. The goal is to reduce debugging time by providing clearer guidance to users when these errors occur.

AI Summary: This issue requests the addition of auto-instrumentation support for Together AI, a platform for hosting open-source models. The implementation is expected to be straightforward due to Together AI's use of an OpenAI-compatible API, making it a good candidate for a "good first issue".

AI Summary: This issue proposes the development of a VS Code extension to integrate AgentDBG directly into the editor. Key features include inline trace visualization, breakpoint support, cost display, and a span tree view. The goal is to provide a seamless debugging experience for developers within their preferred IDE.

AI Summary: This issue requests the addition of a demo GIF or video to the project's README file. The goal is to visually showcase AgentDBG's core features like real-time tracing, pausing, stepping, and cost tracking to enhance user understanding and adoption. The suggested approach involves using an existing example script, recording the demo with a screen recording tool, and optimizing it for size and duration.

AI Summary: This issue requests the addition of auto-instrumentation support for Groq, a fast LLM inference provider. The implementation involves creating a new Python file, leveraging Groq's OpenAI-compatible API, and integrating it into the existing auto-instrumentation function, along with adding tests and an example. The issue is marked as a 'good first issue' due to its straightforward nature.

AI Summary: This issue requests the addition of auto-instrumentation support for the Mistral AI LLM API. The goal is to integrate Mistral's models, which are growing in popularity, into the existing system by wrapping client methods, extracting token counts, and adding tests. The implementation guide suggests following the pattern of existing instrumentors, indicating a straightforward process.

AI Summary: This issue requests the addition of auto-instrumentation support for the Cohere LLM API within the agentdbg tool. The goal is to enable developers to debug agents that utilize Cohere by tracking token counts and other relevant information. The implementation involves creating a new instrumentor file, wrapping specific Cohere client methods, and adding tests and examples.